The Site Data Quality dashboard is designed to give transparency on the quality of the raw input data, and indicate when poor quality data has been modified. Metrics include:
Mislabelled Dump/Load Events: Number of dump/load events where the GPS location falls within a stockpile boundary, but the location label indicated in the FMS raw data indicates a different location. This metric measures the confidence in the FMS raw data location labelling only and thus is not included as an Invalid Dump/Load
Null GPS: Number of dump/load events that have null GPS location but the FMS raw data location label indicates it falls within a modelled stockpile
Outside Boundary: Dump/Load events with valid GPS, occurring outside stockpile boundaries indicated by FMS location label
Invalid Dump/Load: The sum of Null NPS and Outside Boundary
GPS Filling: GPS filling is a technique that assigns/reassigns GPS coordinates of Invalid Dump/Load events. For events that the application is reasonably confident happened in a specific location, GPS coordinates are assigned using a neural network to predict where on a stockpile they were deposited. You can find out more in this help article.
This dashboard includes the following widgets:
-The total number of dump/load events within the chosen period of time for the site as a whole.
-Line graphs that displays the trends of the parameters described previously throughout time, comparing the dump events count by mislabelled vs null GPS vs outside boundary. The null GPS and outside boundary are then combined to make the total invalid dump count line graph.
-The bar charts show the count of invalid dumps and loads per stockpiles and the number of events that GPS filling has been applied to.
-The lower line graphs show the trend of each of the data quality parameters throughout time. This allows the user to be able to identify more precisely where any raw data issues may lie e.g. on which stockpile / specific period of time.
Further reading: